Areas of spatial interaction for a hyperacuity stimulus

Crowding alters the spatial distribution of attention modulation in human primary visual cortex

Crowding effect is the visibility reduction of a target when presented with neighboring distractors. It has been explained by either lateral inhibition at a pre-attentive level or coarse spatial resolution of attention. To test these theories, high-resolution fMRI was used to measure V1 response to the target in the presence or the absence of the distractors in both attended and unattended conditions. We found the cortical response to the target was not affected by the presence of distractors in the unattended condition. However, the spatial distribution of attention modulation in the target and its surrounding area depended on the crowding configuration. When distractors were placed in the same radial axis as the target, a configuration with a severe crowding effect, significant attention enhancements were observed not only in the target's and the distractors' locations, but also in regions next to the target where even no stimulus was presented. But this spread of attention enhancement did not occur when distractors were placed in the same circumference as the target, a configuration with a weak crowding effect. The pattern of interaction between attention and target-distractor configuration supports that crowding results from coarse spatial resolution of attention.

Psychophysical end-stopping associated with line targets

Increment threshold for a small (e.g. 1' x 5') line target superimposed on backgrounds of various shapes and sizes was measured to provide a detailed map of the spatial interactions about line targets. This modified "Westheimer paradigm" indicated sensitization in the length direction as well as in the width direction around the line target. The effect of the adaptation field summed over an elongated, end-tapered central region, and showed strong end-zone antagonism beyond the ends of the elongated summation area, as well as flank antagonism to the sides. Secondary disinhibitory and inhibitory areas outside of the antagonistic surround were also demonstrated. When length of the test line was varied, the length of the summation region increased concomitantly, while the length of the end-zones remained fixed. End-zone antagonism was slightly weaker at oblique orientations. These results demonstrate a perceptual analog to neurophysiological end-stopping, and suggest a multilobed y-dimension weighting profile appropriate for models of spatial visual abilities.

Navigational range lights: effect of stimulus configuration on alignment accuracy

The use of range lights in nautical navigation involves the visual determination of the vertical alignment of two points of light, in order to indicate the centre of the channel. In two experiments it is shown that this vernier alignment task is more accurate if illuminated linear stimuli are used as range lights. Experiment 1 involved a laboratory simulation and experiment 2 involved a field test with full sized range lights. The average reduction of error through the use of linear stimuli over the standard point stimuli, for navigational alignments at a distance of one-half nautical mile from the range lights, was 69-4%. This suggests that a marked safety advantage might accrue from the use of linear range lights for nautical navigation purposes.

Binocular processes in vernier acuity

We investigate the role of binocular mechanisms in vernier acuity, using dichoptic variants of spatial-frequency masking and flank-line interference paradigms. The finding that grating masks and flanking lines presented to one eye elevate (worsen) thresholds for detecting vernier offsets presented to the other eye suggests that neural mechanisms mediating vernier acuity receive binocular inputs, thus placing the loci of these mechanisms at postreceptoral sites. The observation that these threshold elevation effects are orientation dependent is consistent with a contribution to vernier acuity from oriented cortical filters.

Vernier acuity for edges defined by flicker

We investigated the effects of contrast and temporal frequency on vernier edge alignment thresholds. Edges were defined by the presence of a 180 deg phase difference in the temporal modulation waveform of adjacent rectangles with spatially uniform luminances. Thresholds of 10 arc sec or less could be obtained at high contrasts, and flicker rates up to 8 Hz. Above this range, thresholds increased rapidly with decreasing contrast and increasing rates of flicker. Thresholds also increased with increasing temporal frequency over the range 0-20 Hz for contrasts normalized to thresholds for the detection of either flicker or the edge. Linear regression on log-log plots of threshold vs contrast at different temporal frequencies showed that the relationship between threshold and contrast was well described by a power law with an exponent of about -0.5 at temporal frequencies of 8 Hz or lower. About 8 Hz the slope more than doubled and thresholds increased. Thresholds also increased when the relative phase (i.e. the instantaneous sign of the contrast) of the upper and lower edges was reversed, and this effect was observed at all temporal frequencies. Measurements of threshold as a function of the size of a gap between the upper and lower edges suggested that the integration region was larger at 16 Hz or above than at 8 Hz. The results suggest that the channels which mediate vernier hyperacuity are phase sensitive and attenuate frequencies higher than 8 Hz.

Spatial localization of motion-defined and luminance-defined contours

Thresholds for the vernier alignment of contours defined by luminance and coherent random-dot motion were measured. The luminance-defined contours were localized with a precision better than the receptor grain, while the motion-defined contours were localized more poorly than this limit. When motion-defined and luminance-defined targets were matched for dot density, vernier thresholds were equivalent at low densities. When the targets were also equated for perceived contrast, the vernier thresholds became equivalent at higher densities as well. These results suggest that the precision with which motion-defined contours are localized is contrast and sample limited. Next, the localization mechanism for motion-defined targets was investigated. Length summation limits were similar for motion-defined and luminance-defined targets, suggesting that these targets could be localized by a common mechanism. Vernier targets were then flanked by two additional bars. Motion-defined flanks interfered with the localization of motion-defined targets and luminance-defined flanks interfered with the localization of luminance-defined targets. However, motion-defined and luminance-defined bars did not interact to produce spatial interference. This result indicates that the mechanisms for localizing luminance-defined and motion-defined targets are independent. We suggest that parallel mechanisms govern the vernier localization of motion-defined and luminance-defined targets.

The two-dimensional shape of spatial interaction zones in the parafovea

The spatial analysis of a target may be strongly degraded by the simultaneous presentation of nearby pattern elements. The present study investigated the shape and extent of the region of interaction as a function of retinal location. The stimuli consisted of 3 collinear [symbol: see text] s which were randomly oriented up ([symbol: see text]) or down ([symbol: see text]). The task was to discriminate the orientation of the middle [symbol: see text]. The retinal locations studied were at 0, 2.5, 5 and 10 degrees, on the lower vertical meridian and on the nasal halves of both the horizontal and the 45 degrees diagonal visual field meridians. The extent of the interaction region was defined as the separation between the midpoint of two adjacent [symbol: see text] s that resulted in 75% correct discrimination. The shape of the interaction region was determined by using several orientations (horizontal, vertical, left diagonal and right diagonal) for the virtual line joining the 3. [symbol: see text] s. Our results show that the size of the interaction regions varies linearly with eccentricity as does the size of a just resolved individual [symbol: see text]. However, the size of the interaction region varies much more rapidly than does the resolution threshold for an individual [symbol: see text]. The spatial interaction zones appear to be elongated radially, so that they have an elliptical shape. The size of the major axis is about 2-3 times the size of the minor axis. The major axis is along the meridian through the central visual field (i.e. it is oriented radially) while the minor axis is oriented tangentially (i.e. isoeccentrically).

Colliding targets: evidence for spatial localization within the motion system

The ability to judge the relative location of moving targets is seriously degraded if the targets move in different directions. The vernier acuity for a target in which the two components move in the same direction is not impaired until target velocity exceeds about 4 deg/sec. If the components are moving along trajectories which differ in direction by more than 15 deg, vernier thresholds rise significantly at target speeds greater than 1 deg/sec. The conditions which affect stationary vernier acuity, i.e. separation of target components, duration, and orientation, do not account for the loss in acuity. Our results suggest that localization for moving targets depends on directionally-selective motion detectors.